bert-base-mnli / trainer_state.json
Cheng98's picture
Upload folder using huggingface_hub
cbf719c verified
raw
history blame contribute delete
No virus
30.6 kB
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 5.0,
"global_step": 122720,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.02,
"learning_rate": 1.991900260756193e-05,
"loss": 0.8698,
"step": 500
},
{
"epoch": 0.04,
"learning_rate": 1.983751629726206e-05,
"loss": 0.7184,
"step": 1000
},
{
"epoch": 0.06,
"learning_rate": 1.9756029986962192e-05,
"loss": 0.6715,
"step": 1500
},
{
"epoch": 0.08,
"learning_rate": 1.9674543676662323e-05,
"loss": 0.6418,
"step": 2000
},
{
"epoch": 0.1,
"learning_rate": 1.9593057366362454e-05,
"loss": 0.6175,
"step": 2500
},
{
"epoch": 0.12,
"learning_rate": 1.9511734028683183e-05,
"loss": 0.6239,
"step": 3000
},
{
"epoch": 0.14,
"learning_rate": 1.9430247718383314e-05,
"loss": 0.5822,
"step": 3500
},
{
"epoch": 0.16,
"learning_rate": 1.9348761408083442e-05,
"loss": 0.5878,
"step": 4000
},
{
"epoch": 0.18,
"learning_rate": 1.9267275097783573e-05,
"loss": 0.5581,
"step": 4500
},
{
"epoch": 0.2,
"learning_rate": 1.9185788787483704e-05,
"loss": 0.5579,
"step": 5000
},
{
"epoch": 0.22,
"learning_rate": 1.9104302477183832e-05,
"loss": 0.5577,
"step": 5500
},
{
"epoch": 0.24,
"learning_rate": 1.9022816166883967e-05,
"loss": 0.5564,
"step": 6000
},
{
"epoch": 0.26,
"learning_rate": 1.8941329856584098e-05,
"loss": 0.5554,
"step": 6500
},
{
"epoch": 0.29,
"learning_rate": 1.8860006518904827e-05,
"loss": 0.5376,
"step": 7000
},
{
"epoch": 0.31,
"learning_rate": 1.8778520208604955e-05,
"loss": 0.5277,
"step": 7500
},
{
"epoch": 0.33,
"learning_rate": 1.8697033898305086e-05,
"loss": 0.5205,
"step": 8000
},
{
"epoch": 0.35,
"learning_rate": 1.8615547588005217e-05,
"loss": 0.5252,
"step": 8500
},
{
"epoch": 0.37,
"learning_rate": 1.8534061277705345e-05,
"loss": 0.5378,
"step": 9000
},
{
"epoch": 0.39,
"learning_rate": 1.8452737940026077e-05,
"loss": 0.5264,
"step": 9500
},
{
"epoch": 0.41,
"learning_rate": 1.837125162972621e-05,
"loss": 0.524,
"step": 10000
},
{
"epoch": 0.43,
"learning_rate": 1.828976531942634e-05,
"loss": 0.5101,
"step": 10500
},
{
"epoch": 0.45,
"learning_rate": 1.8208279009126467e-05,
"loss": 0.5028,
"step": 11000
},
{
"epoch": 0.47,
"learning_rate": 1.81267926988266e-05,
"loss": 0.5145,
"step": 11500
},
{
"epoch": 0.49,
"learning_rate": 1.804546936114733e-05,
"loss": 0.5175,
"step": 12000
},
{
"epoch": 0.51,
"learning_rate": 1.796398305084746e-05,
"loss": 0.5091,
"step": 12500
},
{
"epoch": 0.53,
"learning_rate": 1.788249674054759e-05,
"loss": 0.502,
"step": 13000
},
{
"epoch": 0.55,
"learning_rate": 1.780101043024772e-05,
"loss": 0.5066,
"step": 13500
},
{
"epoch": 0.57,
"learning_rate": 1.7719524119947852e-05,
"loss": 0.5093,
"step": 14000
},
{
"epoch": 0.59,
"learning_rate": 1.763803780964798e-05,
"loss": 0.5021,
"step": 14500
},
{
"epoch": 0.61,
"learning_rate": 1.755655149934811e-05,
"loss": 0.5129,
"step": 15000
},
{
"epoch": 0.63,
"learning_rate": 1.747522816166884e-05,
"loss": 0.4993,
"step": 15500
},
{
"epoch": 0.65,
"learning_rate": 1.739374185136897e-05,
"loss": 0.4913,
"step": 16000
},
{
"epoch": 0.67,
"learning_rate": 1.7312255541069102e-05,
"loss": 0.4961,
"step": 16500
},
{
"epoch": 0.69,
"learning_rate": 1.7230769230769234e-05,
"loss": 0.4865,
"step": 17000
},
{
"epoch": 0.71,
"learning_rate": 1.7149282920469365e-05,
"loss": 0.5059,
"step": 17500
},
{
"epoch": 0.73,
"learning_rate": 1.7067796610169493e-05,
"loss": 0.5019,
"step": 18000
},
{
"epoch": 0.75,
"learning_rate": 1.6986310299869624e-05,
"loss": 0.4809,
"step": 18500
},
{
"epoch": 0.77,
"learning_rate": 1.6904823989569755e-05,
"loss": 0.484,
"step": 19000
},
{
"epoch": 0.79,
"learning_rate": 1.6823500651890484e-05,
"loss": 0.4843,
"step": 19500
},
{
"epoch": 0.81,
"learning_rate": 1.6742014341590612e-05,
"loss": 0.4647,
"step": 20000
},
{
"epoch": 0.84,
"learning_rate": 1.6660528031290743e-05,
"loss": 0.4703,
"step": 20500
},
{
"epoch": 0.86,
"learning_rate": 1.6579204693611475e-05,
"loss": 0.4732,
"step": 21000
},
{
"epoch": 0.88,
"learning_rate": 1.6497718383311606e-05,
"loss": 0.4914,
"step": 21500
},
{
"epoch": 0.9,
"learning_rate": 1.6416232073011734e-05,
"loss": 0.4647,
"step": 22000
},
{
"epoch": 0.92,
"learning_rate": 1.6334745762711865e-05,
"loss": 0.4749,
"step": 22500
},
{
"epoch": 0.94,
"learning_rate": 1.6253259452411997e-05,
"loss": 0.4835,
"step": 23000
},
{
"epoch": 0.96,
"learning_rate": 1.6171936114732725e-05,
"loss": 0.4741,
"step": 23500
},
{
"epoch": 0.98,
"learning_rate": 1.6090449804432857e-05,
"loss": 0.4649,
"step": 24000
},
{
"epoch": 1.0,
"learning_rate": 1.6008963494132988e-05,
"loss": 0.4664,
"step": 24500
},
{
"epoch": 1.02,
"learning_rate": 1.592747718383312e-05,
"loss": 0.3638,
"step": 25000
},
{
"epoch": 1.04,
"learning_rate": 1.5845990873533247e-05,
"loss": 0.3646,
"step": 25500
},
{
"epoch": 1.06,
"learning_rate": 1.5764504563233378e-05,
"loss": 0.3599,
"step": 26000
},
{
"epoch": 1.08,
"learning_rate": 1.568301825293351e-05,
"loss": 0.3495,
"step": 26500
},
{
"epoch": 1.1,
"learning_rate": 1.5601694915254238e-05,
"loss": 0.3495,
"step": 27000
},
{
"epoch": 1.12,
"learning_rate": 1.552020860495437e-05,
"loss": 0.3526,
"step": 27500
},
{
"epoch": 1.14,
"learning_rate": 1.54387222946545e-05,
"loss": 0.3608,
"step": 28000
},
{
"epoch": 1.16,
"learning_rate": 1.5357235984354632e-05,
"loss": 0.349,
"step": 28500
},
{
"epoch": 1.18,
"learning_rate": 1.527591264667536e-05,
"loss": 0.3538,
"step": 29000
},
{
"epoch": 1.2,
"learning_rate": 1.519442633637549e-05,
"loss": 0.3482,
"step": 29500
},
{
"epoch": 1.22,
"learning_rate": 1.511294002607562e-05,
"loss": 0.3652,
"step": 30000
},
{
"epoch": 1.24,
"learning_rate": 1.503145371577575e-05,
"loss": 0.3665,
"step": 30500
},
{
"epoch": 1.26,
"learning_rate": 1.494996740547588e-05,
"loss": 0.3628,
"step": 31000
},
{
"epoch": 1.28,
"learning_rate": 1.4868644067796611e-05,
"loss": 0.356,
"step": 31500
},
{
"epoch": 1.3,
"learning_rate": 1.4787157757496742e-05,
"loss": 0.3613,
"step": 32000
},
{
"epoch": 1.32,
"learning_rate": 1.4705834419817473e-05,
"loss": 0.3662,
"step": 32500
},
{
"epoch": 1.34,
"learning_rate": 1.4624348109517604e-05,
"loss": 0.3461,
"step": 33000
},
{
"epoch": 1.36,
"learning_rate": 1.4542861799217733e-05,
"loss": 0.3552,
"step": 33500
},
{
"epoch": 1.39,
"learning_rate": 1.4461375488917863e-05,
"loss": 0.3655,
"step": 34000
},
{
"epoch": 1.41,
"learning_rate": 1.4379889178617994e-05,
"loss": 0.3433,
"step": 34500
},
{
"epoch": 1.43,
"learning_rate": 1.4298402868318124e-05,
"loss": 0.3578,
"step": 35000
},
{
"epoch": 1.45,
"learning_rate": 1.4216916558018255e-05,
"loss": 0.3497,
"step": 35500
},
{
"epoch": 1.47,
"learning_rate": 1.4135430247718384e-05,
"loss": 0.3614,
"step": 36000
},
{
"epoch": 1.49,
"learning_rate": 1.4053943937418515e-05,
"loss": 0.3488,
"step": 36500
},
{
"epoch": 1.51,
"learning_rate": 1.3972457627118645e-05,
"loss": 0.3531,
"step": 37000
},
{
"epoch": 1.53,
"learning_rate": 1.3890971316818776e-05,
"loss": 0.3595,
"step": 37500
},
{
"epoch": 1.55,
"learning_rate": 1.3809647979139507e-05,
"loss": 0.353,
"step": 38000
},
{
"epoch": 1.57,
"learning_rate": 1.3728161668839636e-05,
"loss": 0.3523,
"step": 38500
},
{
"epoch": 1.59,
"learning_rate": 1.3646675358539767e-05,
"loss": 0.3652,
"step": 39000
},
{
"epoch": 1.61,
"learning_rate": 1.3565189048239897e-05,
"loss": 0.3504,
"step": 39500
},
{
"epoch": 1.63,
"learning_rate": 1.3483865710560626e-05,
"loss": 0.355,
"step": 40000
},
{
"epoch": 1.65,
"learning_rate": 1.3402379400260757e-05,
"loss": 0.3391,
"step": 40500
},
{
"epoch": 1.67,
"learning_rate": 1.3320893089960886e-05,
"loss": 0.3525,
"step": 41000
},
{
"epoch": 1.69,
"learning_rate": 1.3239569752281619e-05,
"loss": 0.345,
"step": 41500
},
{
"epoch": 1.71,
"learning_rate": 1.3158083441981748e-05,
"loss": 0.3582,
"step": 42000
},
{
"epoch": 1.73,
"learning_rate": 1.3076597131681878e-05,
"loss": 0.3578,
"step": 42500
},
{
"epoch": 1.75,
"learning_rate": 1.2995110821382009e-05,
"loss": 0.3525,
"step": 43000
},
{
"epoch": 1.77,
"learning_rate": 1.2913624511082138e-05,
"loss": 0.3504,
"step": 43500
},
{
"epoch": 1.79,
"learning_rate": 1.283213820078227e-05,
"loss": 0.3531,
"step": 44000
},
{
"epoch": 1.81,
"learning_rate": 1.27506518904824e-05,
"loss": 0.3661,
"step": 44500
},
{
"epoch": 1.83,
"learning_rate": 1.266916558018253e-05,
"loss": 0.3519,
"step": 45000
},
{
"epoch": 1.85,
"learning_rate": 1.258767926988266e-05,
"loss": 0.3459,
"step": 45500
},
{
"epoch": 1.87,
"learning_rate": 1.250635593220339e-05,
"loss": 0.3635,
"step": 46000
},
{
"epoch": 1.89,
"learning_rate": 1.242503259452412e-05,
"loss": 0.3617,
"step": 46500
},
{
"epoch": 1.91,
"learning_rate": 1.234354628422425e-05,
"loss": 0.3668,
"step": 47000
},
{
"epoch": 1.94,
"learning_rate": 1.226205997392438e-05,
"loss": 0.3416,
"step": 47500
},
{
"epoch": 1.96,
"learning_rate": 1.2180573663624513e-05,
"loss": 0.3568,
"step": 48000
},
{
"epoch": 1.98,
"learning_rate": 1.2099087353324644e-05,
"loss": 0.3688,
"step": 48500
},
{
"epoch": 2.0,
"learning_rate": 1.2017601043024774e-05,
"loss": 0.3383,
"step": 49000
},
{
"epoch": 2.02,
"learning_rate": 1.1936114732724903e-05,
"loss": 0.2534,
"step": 49500
},
{
"epoch": 2.04,
"learning_rate": 1.1854628422425034e-05,
"loss": 0.242,
"step": 50000
},
{
"epoch": 2.06,
"learning_rate": 1.1773305084745763e-05,
"loss": 0.2292,
"step": 50500
},
{
"epoch": 2.08,
"learning_rate": 1.1691981747066494e-05,
"loss": 0.2383,
"step": 51000
},
{
"epoch": 2.1,
"learning_rate": 1.1610495436766625e-05,
"loss": 0.2377,
"step": 51500
},
{
"epoch": 2.12,
"learning_rate": 1.1529009126466754e-05,
"loss": 0.2391,
"step": 52000
},
{
"epoch": 2.14,
"learning_rate": 1.1447685788787483e-05,
"loss": 0.235,
"step": 52500
},
{
"epoch": 2.16,
"learning_rate": 1.1366199478487615e-05,
"loss": 0.2349,
"step": 53000
},
{
"epoch": 2.18,
"learning_rate": 1.1284713168187746e-05,
"loss": 0.2543,
"step": 53500
},
{
"epoch": 2.2,
"learning_rate": 1.1203226857887877e-05,
"loss": 0.24,
"step": 54000
},
{
"epoch": 2.22,
"learning_rate": 1.1121740547588006e-05,
"loss": 0.2421,
"step": 54500
},
{
"epoch": 2.24,
"learning_rate": 1.1040254237288138e-05,
"loss": 0.2528,
"step": 55000
},
{
"epoch": 2.26,
"learning_rate": 1.0958767926988267e-05,
"loss": 0.2447,
"step": 55500
},
{
"epoch": 2.28,
"learning_rate": 1.0877281616688398e-05,
"loss": 0.2375,
"step": 56000
},
{
"epoch": 2.3,
"learning_rate": 1.0795795306388528e-05,
"loss": 0.2414,
"step": 56500
},
{
"epoch": 2.32,
"learning_rate": 1.0714308996088657e-05,
"loss": 0.2362,
"step": 57000
},
{
"epoch": 2.34,
"learning_rate": 1.0632822685788789e-05,
"loss": 0.2319,
"step": 57500
},
{
"epoch": 2.36,
"learning_rate": 1.0551499348109517e-05,
"loss": 0.2485,
"step": 58000
},
{
"epoch": 2.38,
"learning_rate": 1.047001303780965e-05,
"loss": 0.238,
"step": 58500
},
{
"epoch": 2.4,
"learning_rate": 1.038852672750978e-05,
"loss": 0.2423,
"step": 59000
},
{
"epoch": 2.42,
"learning_rate": 1.0307040417209911e-05,
"loss": 0.2451,
"step": 59500
},
{
"epoch": 2.44,
"learning_rate": 1.022555410691004e-05,
"loss": 0.2468,
"step": 60000
},
{
"epoch": 2.46,
"learning_rate": 1.014423076923077e-05,
"loss": 0.238,
"step": 60500
},
{
"epoch": 2.49,
"learning_rate": 1.00627444589309e-05,
"loss": 0.2414,
"step": 61000
},
{
"epoch": 2.51,
"learning_rate": 9.981258148631032e-06,
"loss": 0.2559,
"step": 61500
},
{
"epoch": 2.53,
"learning_rate": 9.899771838331161e-06,
"loss": 0.2489,
"step": 62000
},
{
"epoch": 2.55,
"learning_rate": 9.818285528031292e-06,
"loss": 0.2485,
"step": 62500
},
{
"epoch": 2.57,
"learning_rate": 9.736799217731422e-06,
"loss": 0.2477,
"step": 63000
},
{
"epoch": 2.59,
"learning_rate": 9.655312907431551e-06,
"loss": 0.2515,
"step": 63500
},
{
"epoch": 2.61,
"learning_rate": 9.573826597131683e-06,
"loss": 0.2442,
"step": 64000
},
{
"epoch": 2.63,
"learning_rate": 9.492503259452413e-06,
"loss": 0.234,
"step": 64500
},
{
"epoch": 2.65,
"learning_rate": 9.411179921773142e-06,
"loss": 0.2615,
"step": 65000
},
{
"epoch": 2.67,
"learning_rate": 9.329693611473273e-06,
"loss": 0.2369,
"step": 65500
},
{
"epoch": 2.69,
"learning_rate": 9.248207301173405e-06,
"loss": 0.247,
"step": 66000
},
{
"epoch": 2.71,
"learning_rate": 9.166720990873534e-06,
"loss": 0.2511,
"step": 66500
},
{
"epoch": 2.73,
"learning_rate": 9.085397653194265e-06,
"loss": 0.2351,
"step": 67000
},
{
"epoch": 2.75,
"learning_rate": 9.003911342894394e-06,
"loss": 0.2503,
"step": 67500
},
{
"epoch": 2.77,
"learning_rate": 8.922425032594525e-06,
"loss": 0.2464,
"step": 68000
},
{
"epoch": 2.79,
"learning_rate": 8.840938722294655e-06,
"loss": 0.2415,
"step": 68500
},
{
"epoch": 2.81,
"learning_rate": 8.759452411994786e-06,
"loss": 0.2564,
"step": 69000
},
{
"epoch": 2.83,
"learning_rate": 8.678129074315515e-06,
"loss": 0.2434,
"step": 69500
},
{
"epoch": 2.85,
"learning_rate": 8.596642764015646e-06,
"loss": 0.2448,
"step": 70000
},
{
"epoch": 2.87,
"learning_rate": 8.515156453715776e-06,
"loss": 0.2488,
"step": 70500
},
{
"epoch": 2.89,
"learning_rate": 8.433670143415907e-06,
"loss": 0.2349,
"step": 71000
},
{
"epoch": 2.91,
"learning_rate": 8.352183833116038e-06,
"loss": 0.2553,
"step": 71500
},
{
"epoch": 2.93,
"learning_rate": 8.270697522816167e-06,
"loss": 0.2416,
"step": 72000
},
{
"epoch": 2.95,
"learning_rate": 8.189211212516299e-06,
"loss": 0.2438,
"step": 72500
},
{
"epoch": 2.97,
"learning_rate": 8.107724902216428e-06,
"loss": 0.2478,
"step": 73000
},
{
"epoch": 2.99,
"learning_rate": 8.026401564537159e-06,
"loss": 0.2412,
"step": 73500
},
{
"epoch": 3.01,
"learning_rate": 7.944915254237288e-06,
"loss": 0.1833,
"step": 74000
},
{
"epoch": 3.04,
"learning_rate": 7.86342894393742e-06,
"loss": 0.1688,
"step": 74500
},
{
"epoch": 3.06,
"learning_rate": 7.781942633637549e-06,
"loss": 0.1683,
"step": 75000
},
{
"epoch": 3.08,
"learning_rate": 7.70045632333768e-06,
"loss": 0.1695,
"step": 75500
},
{
"epoch": 3.1,
"learning_rate": 7.61913298565841e-06,
"loss": 0.1667,
"step": 76000
},
{
"epoch": 3.12,
"learning_rate": 7.53780964797914e-06,
"loss": 0.1758,
"step": 76500
},
{
"epoch": 3.14,
"learning_rate": 7.456323337679271e-06,
"loss": 0.1734,
"step": 77000
},
{
"epoch": 3.16,
"learning_rate": 7.374837027379401e-06,
"loss": 0.174,
"step": 77500
},
{
"epoch": 3.18,
"learning_rate": 7.2933507170795315e-06,
"loss": 0.1824,
"step": 78000
},
{
"epoch": 3.2,
"learning_rate": 7.212027379400261e-06,
"loss": 0.1622,
"step": 78500
},
{
"epoch": 3.22,
"learning_rate": 7.130541069100392e-06,
"loss": 0.1887,
"step": 79000
},
{
"epoch": 3.24,
"learning_rate": 7.049054758800522e-06,
"loss": 0.1878,
"step": 79500
},
{
"epoch": 3.26,
"learning_rate": 6.967568448500652e-06,
"loss": 0.1722,
"step": 80000
},
{
"epoch": 3.28,
"learning_rate": 6.886082138200783e-06,
"loss": 0.1736,
"step": 80500
},
{
"epoch": 3.3,
"learning_rate": 6.804595827900913e-06,
"loss": 0.1866,
"step": 81000
},
{
"epoch": 3.32,
"learning_rate": 6.723109517601043e-06,
"loss": 0.1774,
"step": 81500
},
{
"epoch": 3.34,
"learning_rate": 6.6416232073011745e-06,
"loss": 0.1675,
"step": 82000
},
{
"epoch": 3.36,
"learning_rate": 6.560136897001305e-06,
"loss": 0.1875,
"step": 82500
},
{
"epoch": 3.38,
"learning_rate": 6.478650586701435e-06,
"loss": 0.1884,
"step": 83000
},
{
"epoch": 3.4,
"learning_rate": 6.397164276401565e-06,
"loss": 0.1846,
"step": 83500
},
{
"epoch": 3.42,
"learning_rate": 6.315677966101695e-06,
"loss": 0.1778,
"step": 84000
},
{
"epoch": 3.44,
"learning_rate": 6.234354628422426e-06,
"loss": 0.1692,
"step": 84500
},
{
"epoch": 3.46,
"learning_rate": 6.152868318122556e-06,
"loss": 0.1748,
"step": 85000
},
{
"epoch": 3.48,
"learning_rate": 6.071382007822686e-06,
"loss": 0.172,
"step": 85500
},
{
"epoch": 3.5,
"learning_rate": 5.989895697522816e-06,
"loss": 0.1856,
"step": 86000
},
{
"epoch": 3.52,
"learning_rate": 5.908572359843546e-06,
"loss": 0.1588,
"step": 86500
},
{
"epoch": 3.54,
"learning_rate": 5.827086049543678e-06,
"loss": 0.1923,
"step": 87000
},
{
"epoch": 3.57,
"learning_rate": 5.745599739243808e-06,
"loss": 0.1648,
"step": 87500
},
{
"epoch": 3.59,
"learning_rate": 5.664113428943938e-06,
"loss": 0.162,
"step": 88000
},
{
"epoch": 3.61,
"learning_rate": 5.582790091264668e-06,
"loss": 0.1809,
"step": 88500
},
{
"epoch": 3.63,
"learning_rate": 5.501466753585399e-06,
"loss": 0.1782,
"step": 89000
},
{
"epoch": 3.65,
"learning_rate": 5.419980443285528e-06,
"loss": 0.176,
"step": 89500
},
{
"epoch": 3.67,
"learning_rate": 5.3384941329856585e-06,
"loss": 0.1841,
"step": 90000
},
{
"epoch": 3.69,
"learning_rate": 5.257007822685789e-06,
"loss": 0.1849,
"step": 90500
},
{
"epoch": 3.71,
"learning_rate": 5.175521512385919e-06,
"loss": 0.1708,
"step": 91000
},
{
"epoch": 3.73,
"learning_rate": 5.0940352020860495e-06,
"loss": 0.1832,
"step": 91500
},
{
"epoch": 3.75,
"learning_rate": 5.012548891786181e-06,
"loss": 0.1765,
"step": 92000
},
{
"epoch": 3.77,
"learning_rate": 4.93106258148631e-06,
"loss": 0.1833,
"step": 92500
},
{
"epoch": 3.79,
"learning_rate": 4.8495762711864405e-06,
"loss": 0.1775,
"step": 93000
},
{
"epoch": 3.81,
"learning_rate": 4.768252933507171e-06,
"loss": 0.1919,
"step": 93500
},
{
"epoch": 3.83,
"learning_rate": 4.6867666232073015e-06,
"loss": 0.1739,
"step": 94000
},
{
"epoch": 3.85,
"learning_rate": 4.605280312907432e-06,
"loss": 0.172,
"step": 94500
},
{
"epoch": 3.87,
"learning_rate": 4.523794002607562e-06,
"loss": 0.181,
"step": 95000
},
{
"epoch": 3.89,
"learning_rate": 4.442470664928292e-06,
"loss": 0.1637,
"step": 95500
},
{
"epoch": 3.91,
"learning_rate": 4.360984354628422e-06,
"loss": 0.1885,
"step": 96000
},
{
"epoch": 3.93,
"learning_rate": 4.2794980443285535e-06,
"loss": 0.1902,
"step": 96500
},
{
"epoch": 3.95,
"learning_rate": 4.198011734028684e-06,
"loss": 0.1764,
"step": 97000
},
{
"epoch": 3.97,
"learning_rate": 4.1166883963494135e-06,
"loss": 0.1653,
"step": 97500
},
{
"epoch": 3.99,
"learning_rate": 4.035202086049544e-06,
"loss": 0.1733,
"step": 98000
},
{
"epoch": 4.01,
"learning_rate": 3.953715775749674e-06,
"loss": 0.135,
"step": 98500
},
{
"epoch": 4.03,
"learning_rate": 3.872392438070404e-06,
"loss": 0.1212,
"step": 99000
},
{
"epoch": 4.05,
"learning_rate": 3.790906127770535e-06,
"loss": 0.1349,
"step": 99500
},
{
"epoch": 4.07,
"learning_rate": 3.709419817470665e-06,
"loss": 0.1152,
"step": 100000
},
{
"epoch": 4.09,
"learning_rate": 3.6279335071707954e-06,
"loss": 0.1307,
"step": 100500
},
{
"epoch": 4.12,
"learning_rate": 3.5464471968709258e-06,
"loss": 0.1244,
"step": 101000
},
{
"epoch": 4.14,
"learning_rate": 3.4649608865710565e-06,
"loss": 0.1292,
"step": 101500
},
{
"epoch": 4.16,
"learning_rate": 3.3836375488917867e-06,
"loss": 0.1246,
"step": 102000
},
{
"epoch": 4.18,
"learning_rate": 3.302151238591917e-06,
"loss": 0.1318,
"step": 102500
},
{
"epoch": 4.2,
"learning_rate": 3.220664928292047e-06,
"loss": 0.1125,
"step": 103000
},
{
"epoch": 4.22,
"learning_rate": 3.1391786179921773e-06,
"loss": 0.1328,
"step": 103500
},
{
"epoch": 4.24,
"learning_rate": 3.057692307692308e-06,
"loss": 0.1391,
"step": 104000
},
{
"epoch": 4.26,
"learning_rate": 2.9763689700130383e-06,
"loss": 0.129,
"step": 104500
},
{
"epoch": 4.28,
"learning_rate": 2.8948826597131686e-06,
"loss": 0.1314,
"step": 105000
},
{
"epoch": 4.3,
"learning_rate": 2.813396349413299e-06,
"loss": 0.1398,
"step": 105500
},
{
"epoch": 4.32,
"learning_rate": 2.731910039113429e-06,
"loss": 0.1177,
"step": 106000
},
{
"epoch": 4.34,
"learning_rate": 2.650586701434159e-06,
"loss": 0.1321,
"step": 106500
},
{
"epoch": 4.36,
"learning_rate": 2.56910039113429e-06,
"loss": 0.1126,
"step": 107000
},
{
"epoch": 4.38,
"learning_rate": 2.48761408083442e-06,
"loss": 0.1313,
"step": 107500
},
{
"epoch": 4.4,
"learning_rate": 2.4061277705345505e-06,
"loss": 0.1346,
"step": 108000
},
{
"epoch": 4.42,
"learning_rate": 2.324641460234681e-06,
"loss": 0.107,
"step": 108500
},
{
"epoch": 4.44,
"learning_rate": 2.243155149934811e-06,
"loss": 0.1301,
"step": 109000
},
{
"epoch": 4.46,
"learning_rate": 2.1616688396349415e-06,
"loss": 0.1159,
"step": 109500
},
{
"epoch": 4.48,
"learning_rate": 2.080182529335072e-06,
"loss": 0.1224,
"step": 110000
},
{
"epoch": 4.5,
"learning_rate": 1.998859191655802e-06,
"loss": 0.1253,
"step": 110500
},
{
"epoch": 4.52,
"learning_rate": 1.917535853976532e-06,
"loss": 0.1163,
"step": 111000
},
{
"epoch": 4.54,
"learning_rate": 1.8360495436766626e-06,
"loss": 0.1129,
"step": 111500
},
{
"epoch": 4.56,
"learning_rate": 1.754563233376793e-06,
"loss": 0.1409,
"step": 112000
},
{
"epoch": 4.58,
"learning_rate": 1.673076923076923e-06,
"loss": 0.1208,
"step": 112500
},
{
"epoch": 4.6,
"learning_rate": 1.5917535853976534e-06,
"loss": 0.1276,
"step": 113000
},
{
"epoch": 4.62,
"learning_rate": 1.5102672750977838e-06,
"loss": 0.1182,
"step": 113500
},
{
"epoch": 4.64,
"learning_rate": 1.4287809647979139e-06,
"loss": 0.1236,
"step": 114000
},
{
"epoch": 4.67,
"learning_rate": 1.3472946544980445e-06,
"loss": 0.1309,
"step": 114500
},
{
"epoch": 4.69,
"learning_rate": 1.2658083441981748e-06,
"loss": 0.1219,
"step": 115000
},
{
"epoch": 4.71,
"learning_rate": 1.1843220338983051e-06,
"loss": 0.1292,
"step": 115500
},
{
"epoch": 4.73,
"learning_rate": 1.1028357235984355e-06,
"loss": 0.1262,
"step": 116000
},
{
"epoch": 4.75,
"learning_rate": 1.0213494132985658e-06,
"loss": 0.1148,
"step": 116500
},
{
"epoch": 4.77,
"learning_rate": 9.400260756192961e-07,
"loss": 0.117,
"step": 117000
},
{
"epoch": 4.79,
"learning_rate": 8.587027379400262e-07,
"loss": 0.128,
"step": 117500
},
{
"epoch": 4.81,
"learning_rate": 7.772164276401566e-07,
"loss": 0.128,
"step": 118000
},
{
"epoch": 4.83,
"learning_rate": 6.95730117340287e-07,
"loss": 0.1416,
"step": 118500
},
{
"epoch": 4.85,
"learning_rate": 6.142438070404173e-07,
"loss": 0.1209,
"step": 119000
},
{
"epoch": 4.87,
"learning_rate": 5.327574967405477e-07,
"loss": 0.134,
"step": 119500
},
{
"epoch": 4.89,
"learning_rate": 4.5127118644067805e-07,
"loss": 0.1199,
"step": 120000
},
{
"epoch": 4.91,
"learning_rate": 3.6978487614080835e-07,
"loss": 0.1169,
"step": 120500
},
{
"epoch": 4.93,
"learning_rate": 2.8829856584093875e-07,
"loss": 0.1204,
"step": 121000
},
{
"epoch": 4.95,
"learning_rate": 2.0697522816166886e-07,
"loss": 0.1298,
"step": 121500
},
{
"epoch": 4.97,
"learning_rate": 1.2548891786179923e-07,
"loss": 0.1317,
"step": 122000
},
{
"epoch": 4.99,
"learning_rate": 4.416558018252933e-08,
"loss": 0.1255,
"step": 122500
},
{
"epoch": 5.0,
"step": 122720,
"total_flos": 1.2915645671077632e+17,
"train_loss": 0.2863727584966936,
"train_runtime": 10219.2193,
"train_samples_per_second": 192.139,
"train_steps_per_second": 12.009
}
],
"max_steps": 122720,
"num_train_epochs": 5,
"total_flos": 1.2915645671077632e+17,
"trial_name": null,
"trial_params": null
}